Semiconductor devices are used in a variety of electronic applications, such as personal computers, cell phones, digital cameras, and other electronic equipment, as examples. Semiconductor devices are typically fabricated by sequentially depositing insulating (or dielectric) layers, conductive layers, and semiconductive layers of material over a semiconductor substrate, and patterning the various layers using lithography to form circuit components and elements thereon.
A transistor is an element that is utilized extensively in semiconductor devices. There may be millions of transistors on a single integrated circuit (IC), for example. A common type of transistor used in semiconductor device fabrication is a metal oxide semiconductor field effect transistor (MOSFET).
Early MOSFET processes used one type of doping to create either positive or negative channel transistors. More recent designs, referred to as complementary MOS (CMOS) devices, use both positive and negative channel devices, e.g., a positive channel metal oxide semiconductor (PMOS) transistor and a negative channel metal oxide semiconductor (NMOS) transistor, in complementary configurations. An NMOS device negatively charges so that the transistor is turned on or off by the movement of electrons, whereas a PMOS devices involves the movement of electron vacancies. While the manufacture of CMOS devices requires more manufacturing steps and more transistors, CMOS devices are advantageous because they utilize less power, and the devices may be made smaller and faster.
The gate dielectric for MOSFET devices has in the past typically comprised silicon dioxide, which has a dielectric constant of about 3.9. However, as devices are scaled down in size, using silicon dioxide for a gate dielectric becomes a problem because of gate leakage current, which can degrade device performance. Therefore, there is a trend in the industry towards the development of the use of high dielectric constant (k) materials for use as the gate dielectric in MOSFET devices. The term “high k materials” as used herein refers to a dielectric material having a dielectric constant of about 4.0 or greater.
High k gate dielectric development has been identified as one of the future challenges in the 2002 edition of International Technology Roadmap for Semiconductors (ITRS), incorporated herein by reference, which identifies the technological challenges and needs facing the semiconductor industry over the next 15 years. For low power logic (for portable electronic applications, for example), it is important to use devices having low leakage current, in order to extend battery life. Gate leakage current must be controlled in low power applications, as well as sub-threshold leakage, junction leakage, and band-to-band tunneling.
However, one problem with using high k materials as gate dielectric materials is lower mobility, which is undesirable. Device performance using high k dielectric materials tends to suffer from trapped charge in the dielectric layer, which deteriorates the mobility, making the drive current lower than in transistors having silicon dioxide gate oxides, thus reducing the speed and performance of transistors having high k gate dielectric materials.
Polysilicon is most often used as a gate material in transistor devices. Another problem of using high k materials as a gate dielectric is that if polysilicon is used as a gate material in conjunction with a high k gate dielectric, poly (polysilicon) depletion can occur between the gate dielectric and the gate. When a CMOS device is operated in an inversion mode, poly depletion causes an increase in the electrical equivalent gate oxide, e.g., by about 4 to 5 Angstroms. It is desirable for the gate capacitance to be relatively high for increased gate control. However, poly depletion decreases the capacitance and lowers the drive current of the CMOS device, which is undesirable.
What is needed in the art is a transistor design that does not suffer from the poly depletion effect.
Another problem with using a high-k dielectric material as the gate dielectric of a CMOS transistor is referred to in the art as a “Fermi-pinning” effect, which occurs at the interface of the gate electrode and gate dielectric material. Fermi-pinning is a problem that occurs in CMOS devices having both polysilicon and metal gates. The Fermi-pinning effect causes a threshold voltage shift and low mobility, due to the increased charge caused by the Fermi-pinning effect. Fermi-pinning causes an assymmetric turn-on threshold voltage Vt for the two transistors of a CMOS device, which is undesirable.
In prior art CMOS transistor designs, shown in FIG. 20, typically SiO2 was used as a gate dielectric material 580, and polysilicon was used as a gate electrode material 582 and 584. A symmetric threshold voltage Vt for the PMOS device and the NMOS device of a prior art CMOS device 500 was easily achieved using SiO2 as a gate dielectric material 580. The manufacturing steps typically comprised providing a workpiece 502, and forming an n well and p well in the PMOS region and NMOS region, respectively. In some designs, isolation regions 508 were formed between the n well and p well. The gate dielectric 580 was formed over the workpiece 502, and the gate electrode material 582/584 comprising a layer of polysilicon was formed over the gate dielectric 580. The gate electrodes 582 and 584 and gate dielectric 580 were patterned, and the workpiece 502 was lightly doped to form source and drain extensions 507. Spacers 534 were formed over the sidewalls of the gate electrodes 582 and 584 and gate dielectric 580. The workpiece 502 was then more deeply implanted with dopants to form the source and drain regions S and D of the device 500. For the PMOS device, the gate electrode 582 was p-type, which was achieved by doping the polysilicon gate with boron (B). For the NMOS device, the gate electrode 584 was n-type, which was typically achieved by doping the polysilicon gate with phosphorus (P) or arsenic (As), as examples.
In electronics, the “work function” is the energy, usually measured in electron volts, needed to remove an electron from the Fermi level to a point an infinite distance away outside the surface. Work function is a material property of any material, whether the material is a conductor, semiconductor, or dielectric. The work function of a semiconductor can be changed by doping the semiconductor material. For example, undoped polysilicon has a work function of about 4.5 eV, whereas polysilicon doped with boron has a work function of about 5.0 eV. The work function of a semiconductor or conductor directly affects the threshold voltage of a transistor when the material is used as a gate electrode.
In prior art CMOS devices such as device 500 shown in FIG. 20 utilizing SiO2 or SiON as the gate dielectric material 580, the work function of the CMOS device 500 could be changed or tuned by doping the polysilicon used for the gate electrode material 582/584. However, high-k gate dielectric materials such as hafnium-based dielectric materials exhibit Fermi-pinning, which is caused by the interaction of the high-k gate dielectric materials with adjacent materials, e.g., the gate material. When used as a gate dielectric, high k gate dielectric materials pin or fix the work function, so that doping the polysilicon gate material does not change the work function. Thus, a symmetric Vt for the NMOS and PMOS transistors of a CMOS device having a high k material for the gate dielectric cannot be achieved by doping polysilicon gate material, as in SiO2 gate dielectric CMOS devices. A CMOS device having a high k dielectric material for the NMOS and PMOS transistors has an asymmetric Vtn, and Vtp, due to the Fermi-pinning effect of the high k dielectric material. Efforts have been made to improve the quality of high-k dielectric films and resolve the Fermi-pinning problems, but the efforts have resulted in little success.
Thus, what is needed in the art is a CMOS transistor device design and method of manufacturing thereof that has a high-k gate dielectric material and a symmetric Vt for the p channel metal oxide semiconductor (PMOS) and n channel metal oxide semiconductor (NMOS) transistors of the CMOS device, that does not exhibit a polysilicon depletion effect.